Pathological OCT Retinal Layer Segmentation using Branch Residual U-shape Networks

2017/07/16 by Stefanos Apostolopoulos, Apostolopoulos, Stefanos, Sandro De Zanet +8
Computer Science · Medicine · #AI in cancer detection #Computer Vision and Pattern Recognition (cs.CV) #Digital Imaging for Blood Diseases #FOS: Computer and information sciences #Retinal Imaging and Analysis #cs.CV

paper · pdf · doi:10.48550/arxiv.1707.04931

9 pages, 5 figures, MICCAI 2017

arxiv created 2017/07/16 · openalex publication_date 2017/07/16 · arxiv updated 2017/07/18 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/04

Abstract

The automatic segmentation of retinal layer structures enables clinically-relevant quantification and monitoring of eye disorders over time in OCT imaging. Eyes with late-stage diseases are particularly challenging to segment, as their shape is highly warped due to pathological biomarkers. In this context, we propose a novel fully Convolutional Neural Network (CNN) architecture which combines dilated residual blocks in an asymmetric U-shape configuration, and can segment multiple layers of highly pathological eyes in one shot. We validate our approach on a dataset of late-stage AMD patients and demonstrate lower computational costs and higher performance compared to other state-of-the-art methods.

Related

• A deep learning framework for segmentation of retinal layers from OCT images
• Topology guaranteed segmentation of the human retina from OCT using convolutional neural networks
• SD-RetinaNet: Topologically Constrained Semi-Supervised Retinal Lesion and Layer Segmentation in OCT
• Segmentation of Retinal Low-Cost Optical Coherence Tomography Images\n using Deep Learning
• Retinal Image Segmentation with Small Datasets